Vol. 124: 215–222, 2017 DISEASES OF AQUATIC ORGANISMS Published May 11 https://doi.org/10.3354/dao03123 Dis Aquat Org

Characterization of fischeri strains associated with disease outbreak in brill Scophthalmus rhombus

Jose R. López*, Laura Lorenzo, Rafael Alcantara, J. I. Navas

IFAPA Centro Agua del Pino, Junta de Andalucía, Carretera El Rompido-Punta Umbría km 3.8, CP21450 Cartaya, Huelva, Spain

ABSTRACT: Three bacterial isolates were recovered from a disease outbreak with high mortality affecting brill Scophthalmus rhombus (Linnaeus, 1758). Moribund fish showed no external signs of disease, but plentiful haemorrhages were observed in liver. On the basis of phenotypic and genotypic characterization, the isolates were identified as . The phenotypic pro- file of the isolates was basically similar to that of the type strain of this species, although some dis- crepancies were observed, mainly in the BIOLOG GN profile. The main cellular fatty acids of strain a591 were also consistent with this species. The highest 16S rDNA sequence similarities were recorded with the type strain of A. fischeri (99.07%); other Aliivibrio species showed similar- ity values below 96%. The highest sequence similarities with gyrB, rpoD and recA genes were also recorded with A. fischeri type strain (99.31, 98.99 and 95.29% similarity, respectively). DNA−DNA hybridization assays confirmed that these isolates belong to A. fischeri; levels of DNA relatedness were 73.5 to 86.2% with isolate a591 (reciprocal values of 86.9 to 99.04%). Finally, a virulence evaluation of the isolates using Senegalese sole fry was also performed; significant mortalities (100% mortality within 5 d) were recorded by intraperitoneal injection, but only with high doses of (2 × 106 cfu g−1 body weight).

KEY WORDS: Fish pathogen · Diagnosis · Identification · Virulence

Resale or republication not permitted without written consent of the publisher

INTRODUCTION Ruby et al. 2005). These attributes make A. fischeri a useful for examination of microbial The genus Aliivibrio (: - , and bacterial− naceae) was created by Urbanczyk et al. (2007) after animal (Dunn 2012). However, A. fischeri reclassification of a number of Vibrio species from the strains have also been found associated with disease former Vibrio fischeri clade, based on phylogenetic outbreaks in shrimp Penaeus monodon (Fabricius, and phenotypic differences. At present, the genus 1798) (Lavilla-Pitogo et al. 1998) and in marine fish in comprises 6 species: A. fischeri, A. logei, A. salmoni- Spain—namely in sea bream Sparus aurata Lin- cida, A. wodanis, A. finisterrensis and A. sifiae. naeus, 1758 and turbot Scophthalmus maximus Lin- A. fischeri, the type species of the genus, can be naeus, 1758 cultures — although its pathogenicity is found either free living in the marine environment or not clear (Lamas et al. 1990, Balebona et al. 1998, associated with a eukaryotic host; in fact, certain bio- Buller 2004). luminescent strains are symbiont in the light-emit- To our knowledge, the present paper reports the ting organs of certain and fishes (Farmer 2006, first description and characterization of A. fischeri

*Corresponding author: [email protected] © Inter-Research 2017 · www.int-res.com 216 Dis Aquat Org 124: 215–222, 2017

strains associated with a disease outbreak in brill using the MIDI operating system and the aerobic Scophthalmus rhombus. bacteria library TSBA6 (MIDI 2008).

MATERIALS AND METHODS Phylogenetic analysis

Bacterial isolation Template DNA from pure cultures was prepared by boiling bacterial colonies for 5 min in distilled Mortalities occurred in adult-sized fish (1.7 to water followed by centrifugation at 16 168 × g for 2.8 kg) stocked at 3.5 kg m−2 in a recirculating system 2 min to sediment the cell debris. The concentration with fiberglass tanks. Samples for bacterial isolation and purity of genomic DNA was calculated from were taken from liver and kidney of moribund fish, measurements of absorbance at 260 (A260) and and cultured on Flexibacter maritimus medium 280 nm (A280), recorded using a NanoDrop 1000 (FMM) (Pazos et al. 1996) at 20°C for 24 to 96 h. For spectro photometer. Partial 16S rRNA gene se- long-term preservation, strains were frozen at −80°C quences were obtained using the universal primers in sterile seawater supplemented with 20% (v/v) 20F and 1500R, capable of amplifying nearly full- glycerol. length 16S rDNA (Weisburg et al. 1991). Sequencing of the housekeeping genes gyrB, rpoD and recA was performed using the primers proposed by Santos & Phenotypic characterization Ochman (2004), Yamamoto & Harayama (1998) and Islam et al. (2013), respectively. PCRs were per- Phenotypic characterization was performed ac- formed basically as indicated in the literature in each cording to Bernardet et al. (1990) and Avendaño- case in a total reaction volume of 25 µl, using the Herrera et al. (2004). The Gram reaction was deter- commercial kit MyTaqTM DNA Polymerase (Bioline) mined according to the KOH method proposed by which includes all necessary reagents except the Buck (1982) and by the Gram-staining method. primers and DNA. PCR products were purified with Temperature tolerance was tested by checking the commercial kit Illustra ExoProStar 1-step (GE growth on Marine Agar (Difco) at 4, 15, 25, 30, 35, Healthcare) following the manufacturer’s instruc- 40 and 45°C for 10 d, and tolerance of salinity was tions. Direct sequencing of purified PCR products tested with growth on basal medium (neopeptone was performed by Secugen (Madrid). The sequences 4g l−1; yeast extract 1 g l−1; agar 15 g l−1) supple- were analyzed using Chromas LITE and BioEdit pro- mented with 0, 3, 6, 8, 10 and 12% (w/v) NaCl. grams and subjected to BLAST (https://blast.ncbi. Growth on thiosulfate–citrate–bile salts–sucrose nlm.nih.gov) and EzTaxon (www.ezbiocloud.net/ (TCBS) agar (Difco) was also tested. All tests were eztaxon) searches to retrieve the most closely related incubated aerobically at 20°C. Commercial minia- sequences. Se quence similarities were calculated turized API 20E and API 20NE galleries (bio- using SIAS software (http://imed.med.ucm.es/Tools/ Merieux), and Biolog GN2 Micro-plates were also sias). DNA sequences were aligned with others from utilized according to the manufacturer’s instruc- related species using Clustal Omega software, and tions, but sterile sea water was used as a diluent phylogenetic trees were constructed according to and 20°C as the incubation temperature. The type the neighbor-joining method (Saitou & Nei 1987) by strain of Aliivibrio fischeri (CECT 524T) was charac- using the program MEGA. The accuracy of the terized together with the isolates under study, with resulting tree was measured by bootstrap resampling the same methodology. of 1000 replicates.

Analysis of fatty acid methyl esters Identification by PCR

Preparation of fatty acid methyl esters (FAMEs) A PCR assay using a pair of specific primers target- from strain a591, grown at 20°C on Marine Agar ing the luxA gene of A. fischeri was carried out as plates, was performed according to the instructions previously described (Buller 2004), but using a of the Microbial Identification System (MIDI) as higher annealing temperature (60°C). Template described by Sasser (1990). FAMEs were analyzed by DNA was extracted as mentioned previously and gas chromatography in an Agilent 6850 system, 100 ng of DNA were used for each strain. PCRs were López et al.: Aliivibrio fischeri in diseased brill 217

performed using the commercial kit MyTaqTM DNA length. After recovery by centrifugation (16 168 × g, Polymerase. DNA from the A. fischeri type strain was 2 min), bacteria were washed in phosphate buffered included as a positive control and distilled water as saline (PBS) and finally re-suspended in PBS. Doses negative control. PCR products were electropho- were confirmed with total viable counts after spread- resed on a 2% agarose Tris-borate-EDTA buffer gel ing 0.1 ml volumes of each dose over the surface of stained with SYBR Safe DNA Gel Stain (Invitrogen). duplicate plates of FMM. A control group (chal- A 100 bp DNA ladder H3RTU (Nippon Genetics) was lenged with PBS only) of 10 fish was included in each included as a molecular weight marker. virulence assay. After bacterial challenge, experi- mental and control fish were kept without feeding in 18 l tanks at 18 to 20°C in continually flowing seawa- Molecular fingerprinting ter, and mortalities were recorded daily for a 10 d period. Dead fish were removed and subjected to Template DNA was extracted as mentioned pre - bacteriological examination as previously indicated; viously and 100 ng of DNA were used for each recovered strains were identified by specific PCR. strain. Repetitive extragenic palindromic (REP)-PCR and enterobacteria repetitive intergenic consensus (ERIC) -PCR analysis were performed as previously RESULTS described (Rodríguez et al. 2006) using the commer- cial kit MyTaqTM DNA Polymerase, in a Veriti 96 well Bacterial isolation thermal cycler (Applied Biosystems). PCR products were electrophoresed as mentioned above. Similarity During July 2013, a epizootic outbreak with a high between isolates and the A. fischeri type strain was mortality rate (near 100%) occurred in a marine farm estimated using the Dice similarity coefficient (SD) located in southwestern Spain, affecting brill Scoph- (Dice 1945) with the software DendroUPGMA thalmus rhombus adult cultures in water tempera- (http://genomes.urv.cat/UPGMA). tures of 18 to 20°C. Affected fish showed no external signs of disease. Internally, however, plentiful haem- orrhages were observed, mainly in the liver (Fig. 1). DNA−DNA hybridization Numerous pale yellow colonies appeared in the cul- ture medium from liver and kidney samples, almost For DNA−DNA hybridization assays, DNA was in pure culture. Three isolates (a589, a590 and a591) extracted with the kit NucleoSpin Tissue (Macherey- were selected for identification. Nagel) and the concentration and purity of each sam- ple were determined by measuring the A260 and A260/A280 ratio, respectively. DNA−DNA hybridiza- Phenotypic characterization tion assays were performed by the plate method pro- posed by Ziemke et al. (1998), combining the hydrox- Colonies of the isolates under study were pale yel- yapatite method with non-radioactive detection of low in color, luminescent, not adherent to agar, and released DNA. The hybridization temperature (Tm) consisted of Gram-negative, fermentative rods. All was 60°C. isolates were cytochrome oxidase and catalase posi- tive. Growth was observed at 4 to 35°C, but not at 40°C. All strains grew in 3 to 6% NaCl, but none in Pathogenicity assays 0% or 8 to 12% NaCl. All isolates were able to grow in TCBS, displaying green colonies. The arginine To investigate the pathogenicity of the isolates, dihydrolase test was negative, but all isolates were experimental infections were performed by intra - positive for lysine and ornithine decarboxylase. peritoneal injection of 3 different doses (2 ×104, 2 × Hydrolysis of Tween-20 and Tween-80 were positive, 105, and 2 × 106 cfu g−1 body weight) of isolate a591 in but starch and casein were not hydrolysed. In API Senegalese sole Solea senegalensis (Kaup, 1858) 20E galleries, positive results were recorded for with weights between 4.5 and 5.5 g. Groups of 10 fish nitrate reduction and acid production from glucose, were used for each dose. This assay was performed mannitol, amygdalin; the ornithine decarboxylase in duplicate. Bacteria were growth in FMM at 20°C test was also generally positive. In API 20NE gal- for 24 h, and bacterial concentration was estimated leries, only nitrate reduction, glucose fermentation, by absorbance of bacterial cultures at 600 nm wave- esculin hydrolysis, and the β-galactosidase test gave 218 Dis Aquat Org 124: 215–222, 2017

glucosamine, D-fructose, D-galactose, gentio biose, α-D-glucose, D-mannose, inosine, uridine, thymidine, glycerol and glucose-6-phosphate.

Analysis of FAMEs

The FAME analyses showed that the major cellular fatty acids (>5% of the

total) in strain a591 were iso-C12:0 (8.34%), iso-C 3OH (6.04%), iso- Fig. 1. Red-tinged fluid in the abdomen and haemorrhages in liver, the main 12:0 symptoms observed in diseased brill Scophthalmus rhombus C14:0 (8.03%), iso-C16:0 (13.79%) and sum in features 2 (5.72%), 3 (48.10%) and 8 (9.22%). These results agreed Table 1. Differences in carbon compound utilization between with the FAME profile published for the A. fischeri Aliivibrio fischeri strains, including both the isolates obtained type strain (Yoshizawa et al. 2010). in this work from brill Scophthalmus rhombus (a589, a590 and a591) and the type strain of A. fischeri (CECT 524T), determined using the BIOLOG GN system Phylogenetic analysis Characteristic Positive strains Almost complete 16S rDNA sequences were ob - Dextrin All except a591 tained from strains a589, a590 and a591 and used Maltose a589 for BLAST and EzTaxon homology searches to D-psicose a589, a590 retrieve the most closely related species. Based on Cis-aconitic acid a589 Sebacic acid a591 sequence analysis, the isolates were included in Bromosuccinic acid a589, CECT 524T the genus Aliivibrio, within the family Vibrio na - L-asparagine a589 ceae. The highest sequence similarity for isolate L-aspartic acid a589 a591 (1431 bp) was recorded with A. fischeri L-leucine a589 ATCC 7744T (99.07% similarity), followed by the L-proline a589 Phenyethylamine All except a591 type strains of A. finisterrensis (95.67%), A. sifiae (94.96%) and A. wodanis (94.89%). Sequence sim- ilarities be tween the 3 isolates under study were 100%. The phylogenetic tree derived from these clear positive results. Malate assimilation was also se quences illustrates the position of the isolates generally positive. recovered from brill, clearly grouped with the A. Similar results were displayed by the type strain of fischeri type strain and separated from other Alii- A. fischeri, except for the following tests: ornithine vibrio and Vibrio species (Fig. 2). Phylogenetic decarboxylase (in both tube and API 20E), growth at analysis based on gyrB gene sequences confirmed 4°C and at 35°C, urease (in both API 20E and API the clustering with high bootstrap values of the 3 20NE), Voges-Proskauer (API 20E) and malate as - isolates with A. fische ri (Fig. 3). The closest type similation (API 20NE). strains were that of A. fischeri (99.31% similarity), Data obtained from GN MicroPlates (Biolog) A. wodanis (85.02%), A. salmonicida (84.51%), A. showed differences between the strains character- logei (83.74%), A. sifiae (80.2%) and A. finister- ized in this work. All the brill isolates utilized the rensis (79.17%). Partial rpoD and recA gene se - following carbon sources: N-acetyl-D-glucosamine, quences were also obtained for the 3 isolates. The D-fructose, D-galactose, gentiobiose, α-D-glucose, D- highest sequence similarities for isolate a591 were mannose, inosine, uridine, thymidine, glycerol and obtained with A. fischeri type strain (98.99 and glucose-6-phosphate. Variable results were found for 95.29% similarity, respectively). Se quences from 11 carbon sources (Table 1). All the isolates were other Aliivibrio species were not available in the negative for the remaining 73 tests. The A. fischeri GenBank database. The GenBank/ EMBL/DDBJ type strain can be clearly differentiated from the accession numbers for the sequences obtained in brill isolates with the following tests: N-acetyl-D- this work are LT616977 to LT616988. López et al.: Aliivibrio fischeri in diseased brill 219

Fig. 2. Neighbor-joining phylogenetic tree based on 16S rRNA gene sequences showing the relationships between the isolates under study and the type strains of related members of Aliivibrio and Vibrio. The Tenacibaculum soleae sequence was used as an outgroup. Numbers at the nodes indicate the levels of bootstrap based on 1000 replicates. Sequences from related species were obtained from GenBank database; their accession numbers are indicated before the species name

Fig. 3. Neighbor-joining phylogenetic tree based on gyrB gene sequences showing the relationships between the isolates under study and the type strains of related members of Aliivibrio and Vibrio. The Aeromonas hydrophila sequence was used as an outgroup. Numbers at the nodes indicate the levels of bootstrap based on 1000 replicates. Accession numbers are indicated before the species name 220 Dis Aquat Org 124: 215–222, 2017

Pathogenicity tests

Strain a591 was selected for experimental infec- tion. Mortalities were observed only at the highest of the 3 doses employed (2 ×106 cfu g−1 body weight), and were 100% within the first 5 d after exposure to the pathogen. The inoculated strains were recovered from most of the dead fish. None of the control fish died during the assays.

DISCUSSION

Fig. 4. (A) Repetitive extragenic palindromic (REP)-PCR and This work reports the first isolation of Aliivibrio fis- (B) enterobacteria repetitive intergenic consensus (ERIC)- PCR patterns of Aliivibrio fischeri strains. MW: 100 bp DNA cheri associated with disease in brill Scophthalmus Ladder H3 RTU molecular size marker (Nippon Genetics); rhombus, a candidate species for diversification of lanes 1 to 5: a589, a590, a591, CECT 524 and negative con- marine aquaculture in Spain. The main symptoms trol. Numbers on the left indicate the position of molecular observed in the outbreak, which reached a mortality size marker (in bp) rate of nearly 100%, consisted of haemorrhages in the liver and the presence of haemorragic liquid in the peritoneal cavity; no external alterations were Identification by PCR and molecular fingerprinting observed. A. fischeri has been previously described associated with disease in other cultured fish in All brill isolates showed the expected 428 bp band, Spain. It was recovered by Lamas et al. (1990) from identical to that of the A. fischeri type strain, when turbot affected by skin papillomas and visceral tu - the specific PCR protocol targeting the luxA gene of mors; in some cases, fish also showed similar symp- this species was employed (data not shown). No toms to those observed in brill (liver with petechia, amplification product was detected in the negative red tinged fluid in the abdomen). Bacteria were re - control. covered from both external lesions and from internal In REP-PCR and ERIC-PCR analysis, a unique pro- organs, and mortality was 39% in a year. A. fischeri file was observed for all brill isolates, clearly different has also been described from sea bream (Balebona et from that of A. fischeri type strain (Fig. 4). Dice coef- al. 1998), although in this case neither symptomatol- ficient between the 3 brill isolates was 1 for both ogy nor mortality data were indicated. Other Aliivib- REP-PCR and ERIC-PCR analysis. On the contrary, rio species have also been associated with disease in Dice coefficient between these isolates and the A. fis- fish exhibiting a variety of symptoms. A. wodanis has cheri type strain was 0.333 for REP-PCR and 0.857 for been found in Atlantic salmon Salmo salar affected ERIC-PCR. In ERIC-PCR, however, clear differences by the winter ulcer disease (Lunder et al. 2000), and were observed between the brill isolates and the type A. logei and A. salmonicida strains cause the Hitra strain with regard to the intensity of a number of disease in salmonids (Egidius et al. 1986, Benedikts- bands. dottir et al. 1998). The symptoms observed in brill were quite similar to that observed with this septicemic disease, characterized by the presence of internal DNA−DNA hybridization haemorrhages and large amounts of red-tinged fluid in the peritoneal cavity. DNA-DNA hybridization experiments with A. fis- Identification of the isolates recovered from dis- cheri type strain were done in duplicate and con- eased brill was performed on the basis of phenotypic firmed the results obtained previously. Levels of and genotypic characterization. Analysis of 16S DNA re-association between strain a591 and the type rDNA sequences showed similarity of 99% between strain of A. fischeri were 73.5 to 86.2%. Reciprocal the isolates and the A. fischeri type strain; on the values were 86.9 to 99.04% with the 3 brill isolates. contrary, these values were always below 96% with On the other hand, DNA−DNA hybridization values other Aliivibrio species. These results place A. fisch - between strain a591 and the other isolates under eri as a unique species, with similarity values above study ranged from 91.7 to 100%. the limit of intraspecific variability (98.7%) proposed López et al.: Aliivibrio fischeri in diseased brill 221

by Stackebrandt & Ebers (2006). Phylogenetic Acknowledgements. This study was funded by project analyses based on housekeeping genes (gyrB, rpoD, BONAQUA (0433_BONAQUA_5_E POCTEP). J.R.L. has a postdoctoral grant from IFAPA supported by the European recA) confirmed the clustering of the 3 isolates with Union FEDER program. A. fischeri and their distinction from other known Aliivibrio species (similarity with this species was 99.31, 98.99, and 95.29%, respectively). On the LITERATURE CITED other hand, results from the analysis of REP-PCR Avendaño-Herrera R, Magariños B, López-Romalde S, and ERIC-PCR showed that the isolates recovered Romalde JL, Toranzo AE (2004) Phenotypic characteriza- from brill constitute a homogeneous group, with tion and description of two major O-serotypes in Tena - band profiles clearly different from that of the A. fis- cibaculum maritimum strains from marine fishes. Dis cheri type strain. Aquat Org 58: 1−8 Balebona MC, Zorilla I, Morinigo MA, Borrego JJ (1998) Biochemical and physiological characteristics of Survey of bacterial pathologies affecting farmed gilt- the isolates, and the FAME profile, were found to head sea bream (Sparus aurata L.) in southwestern Spain basically agree with those reported for A. fischeri by from 1990 to 1996. Aquaculture 166: 19−35 other authors (Buller 2004, Yoshizawa et al. 2010) Benediktsdottir E, Helgason S, Sigurjonsdottir H (1998) Vib- rio spp. isolated from salmonids with shallow skin lesions and with that of the type strain of this species, also and reared at low temperature. J Fish Dis 21: 19−28 characterized in this work. Discrepancies were ob - Bernardet JF, Campbell AC, Buswell JA (1990) Flexibacter served for a number of tests, such as ornithine decar- maritimus is the agent of ‘black patch necrosis’ in Dover boxylase, urease, Voges-Proskauer and growth at 4 sole in Scotland. Dis Aquat Org 8: 233−237 and 35°C; however, the largest differences were Buck JD (1982) Nonstaining (KOH) method for determina- tion of Gram reactions of marine bacteria. Appl Environ observed in Biolog carbon source utilization profiles, Microbiol 44: 992−993 since brill isolates could be easily differentiated from Buller NB (2004) Bacteria from fish and other aquatic ani- A. fischeri type strain by 11 tests. mals: a practical identification manual. CABI Publishing, Despite differences in genotypic and phenotypic Wallingford Dice LR (1945) Measures of the amount of ecological associ- profiles, DNA−DNA hybridization experiments con- ation between species. Ecology 26:297−302 firmed the identification of the isolates as A. fisch- Dunn AK (2012) Vibrio fischeri metabolism: symbiosis and eri. Levels of DNA re-association of strain a591 beyond. Adv Microb Physiol 61: 37−68 with the type strain of A. fischeri and with the other Egidius E, Wiik R, Andersen K, Hoff KA, Hjeltnes B (1986) Vibrio salmonicida sp. nov., a new fish pathogen. Int isolates were 73.5 to 86.2% (reciprocal values were J Syst Bacteriol 36: 518−520 86.9 to 99.04%) and 91.7 to 100%, respectively; Farmer JJ III (2006) The Family . In: Dworkin these values are above the threshold value for spe- M, Falkow S, Rosenberg E, Schleifer KH, Stackebrandt E cies delineation of 70% proposed by Wayne et al. (eds) The prokaryotes, Vol 6. Springer, New York, NY, (1987), proving that these strains all belong to a p 495−507 Islam A, Labbate M, Djordjevic SP, Alam M and others unique species. (2013) Indigenous strains from a non- Finally, experimental infection tests carried out by endemic region are pathogenic. Open Biol 3: 120181 intraperitoneal injection clearly demonstrated the Lamas J, Anadon R, Devesa S, Toranzo AE (1990) Visceral pathogenic potential of these isolates; high mortali- neoplasia and epidermal papillomas in cultured turbot Scophthalmus maximus. Dis Aquat Org 8:179−187 ties rates (100%) were observed, but only at the Lavilla-Pitogo CR, Leaño EM, Paner MG (1998) Mortalities 6 −1 highest dose (2 × 10 cfu g body weight), which of pond-cultured juvenile shrimp, Penaeus monodon, could indicate that these strains act as opportunistic associated with dominance of luminescent in the pathogens. In this way, Lamas et al. (1990) suggested rearing environment. Aquaculture 164: 337−349 Lunder T, Forum H, Holstad G, Steigerwalt AG, Mowinckel that although A. fischeri did not play the primary role P, Brenner DJ (2000) Phenotypic and genotypic charac- in the outbreak described in turbot, this bacterium, terization of Vibrio viscosus sp. nov. and Vibrio wodanis together with , probably contributed to sp. nov. isolated from Atlantic salmon (Salmo salar) with the development of fish mortality. On the other hand, ‘winter ulcer’. Int J Syst Evol Microbiol 50: 427−450 the degree of virulence of the isolates obtained from MIDI (2008) Sherlock microbial identification system operat- ing manual, version 6.1. Microbial Identification Inc. brill match with that observed for A. fischeri strains (MIDI), Newark, DE recovered from sea bream by Balebona et al. (1998), Pazos F, Santos Y, Macias AR, Nuñez S, Toranzo AE (1996) 5 which displayed LD50 values that ranged from 7 × 10 Evaluation of media for the successful culture of Flex- to 1.3 × 107 cfu g−1 body weight. ibacter maritimus. J Fish Dis 19: 193−197 Rodríguez JM, López-Romalde S, Beaz R, Alonso MC, Castro In summary, this work represents the first report of D, Romalde JL (2006) Molecular fingerprinting of Vibrio A. fischeri in brill, and allows the number of potential tapetis strains using three PCR-based methods: ERIC- pathogens that affect this species to be extended. PCR, REP-PCR and RAPD. Dis Aquat Org 69: 175−183 222 Dis Aquat Org 124: 215–222, 2017

Ruby EG, Urbanowski M, Campbell J, Dunn A, Faini M, Aliivibrio salmonicida comb. nov. and Aliivibrio wodanis Gunsalus R (2005) Complete sequence of Vibrio comb. nov. Int J Syst Evol Microbiol 57: 2823−2829 fischeri: a symbiotic bacterium with pathogenic con- Wayne LG, Brenner DJ, Colwell RR, Grimont PAD and oth- geners. Proc Natl Acad Sci USA 102:3004−3009 ers (1987) Report of the ad hoc committee on reconcilia- Saitou N, Nei M (1987) The neighbor-joining method: a new tion of approaches to bacterial systematics. Int J Syst method for reconstructing phylogenetic trees. Mol Biol Bacteriol 37: 463−464 Evol 4: 406−425 Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S Santos SR, Ochman H (2004) Identification and phyloge- ribosomal DNA amplification for phylogenetic study. netic sorting of bacterial lineages with universally con- J Bacteriol 173: 697−703 served genes and . Environ Microbiol 6:754−759 Yamamoto S, Harayama S (1998) Phylogenetic relationships Sasser M (1990) Identification of bacteria by gas chromato - of Pseudomonas putida strains deduced from the nucleo- graphy of cellular fatty acids. MIDI Technical Note 101. tide sequences of gyrB, rpoD and 16S rRNA genes. Int J Microbial Identification Inc. (MIDI), Newark, DE Syst Bacteriol 48: 813−819 Stackebrandt E, Ebers J (2006) Taxonomic parameters re- Yoshizawa S, Karatani H, Wada M, Yokota A, Kogure K visited: tarnished gold standards. Microbiol Today 33: (2010) Aliivibrio sifiae sp. nov., luminous marine bacteria 152−155 isolated from seawater. J Gen Appl Microbiol 56: 509−518 Urbanczyk H, Ast JC, Higgins MJ, Carson J, Dunlap PV Ziemke F, Höfle MG, Lalucat J, Rosselló-Mora R (1998) (2007) Reclassification of Vibrio fischeri, Vibrio logei, Reclassification of Shewanella putrefaciens Owen’s Vibrio salmonicida and Vibrio wodanis as Aliivibrio fis- genomic group II as Shewanella baltica sp. nov. Int J Syst cheri gen. nov., comb. nov., Aliivibrio logei comb. nov., Bacteriol 48: 179−186

Editorial responsibility: Andrew Barnes, Submitted: January 5, 2017; Accepted: March 14, 2017 Brisbane, Queensland, Australia Proofs received from author(s): April 26, 2017